hdflex

Provides a forecasting method that maps vast numbers of (scalar-valued) signals of any type into an aggregate density forecast in a time-varying and computationally fast manner. The method proceeds in two steps: First, it transforms a predictive signal into a density forecast. Second, it combines the generated candidate density forecasts into an ultimate density forecast. The methods are explained in detail in Adaemmer et al. (2023) doi:10.2139/ssrn.4342487.